PEER-REVIEWED ARTICLE Jennifer Cripe* and Mary Losikoff Food Protection Trends, Vol 41, No. 2, p. 184–194 © Copyright 2021, International Association for Food Protection U.S. Food and Drug Administration, Center for Food Safety and Applied 2900 100th Street, Suite 309, Des Moines, IA 50322-3855 Nutrition, 5001 Campus Drive, College Park, MD 20740, USA

Presence of Listeria monocytogenes and Sanitation Controls in Cold-Smoked Facilities during FDA Inspections

ABSTRACT the utmost importance and that implementation of Listeria- The U.S. Food and Drug Administration (FDA) conducted specific controls with environmental sampling should be inspections to observe operating practices and obtain done to prevent cold-smoked salmon from being adulterat- data on the prevalence of Listeria monocytogenes in ed by L. monocytogenes. cold-smoked salmon processing facilities. FDA inspected and sampled 15 cold-smoked salmon processing facilities INTRODUCTION for Listeria spp. Environmental sampling showed that 13 The genusListeria consists of several species, but L. mono- facilities had one or fewer L. monocytogenes positive from cytogenes is considered the species of most concern (23). each facility and two facilities had numerous (10 and 11, L. monocytogenes causes the disease listeriosis and has a respectively) L. monocytogenes positives. The two facilities mortality rate of 20% (2). Listeriosis is a foodborne disease with numerous L. monocytogenes positives had gross in- that can cause fever, diarrhea, nausea, and vomiting in sanitation and poor employee practices. One of these facil- healthy adults (14). More severe symptoms like bacteremia ities had difficult-to-clean surfaces and inadequate cleaning (infection in the blood) and meningitis (swelling of the and sanitizing; these likely contributed to cross-contamina- brain) can occur in elderly and immunocompromised people tion, which led to the adulteration of cold-smoked salmon (14). In pregnant women, the bacterium can pass through the finished product by a resident strain of L. monocytogenes. placenta, leading to miscarriages (14). Uncertainty related The inspection results also showed that the three facilities to the relative pathogenicity of L. monocytogenes strains and where there was no detection of Listeria spp. routinely the susceptibility of individual human hosts makes limiting conducted environmental and finished product testing, exposure the best means of reducing infections (22). The frequently trained employees, and did routine cleaning and U.S. Food and Drug Administration (FDA) considers ready- sanitizing. Overall, the results indicate that sanitation is of to-eat food (RTE) contaminated with L. monocytogenes to be

*Author for correspondence: Phone: +1 240.402.3045; Fax: +1 301.436.2632; Email: [email protected]

184 Food Protection Trends March/April adulterated under section 402 (a)(1) of the Federal Food, plant B was 1.2%. Ribotyping of isolates showed that some Drug, and Cosmetic Act. strains from the environmental samples were not detected Listeria monocytogenes can grow with or without oxygen, at in the raw fish samples and that persistentL. monocytogenes a maximum water phase salt of 10%, a minimum pH of 4.4, strains in the plant environment were the primary source of and minimum water activity of 0.92 (21). L. monocytogenes contamination of cold-. The study concluded that can grow at a minimum temperature of −0.4°C (31.2°F) (i) environmental contamination is largely separate from the (21). In a survey of 130 samples from 13 lots of refrigerated, incoming raw fish, (ii) persistent strains within the plant can vacuum-packaged, cold-smoked salmon packed in six Danish be a continual source of contamination, and (iii) insanitation smokehouses, the pH and water phase salt ranged from 6.1 in the plant significantly contributes toL. monocytogenes to 6.3 and 3.5 to 8.9%, respectively (7). These levels provide contamination (11). a suitable environment for the growth of L. monocytogenes. The importance of sanitation andL. monocytogenes control A review of six studies showed an exponential growth rate measures was further supported by a 2004 study of four of 0.050 to 0.556 log CFU/day at 5°C (41°F) for L. mono- cold-smoked salmon processing plants in the United States cytogenes inoculated in cold-smoked salmon (26). These (19). A total of 234 raw fish, 233 finished products, and 533 results indicate that L. monocytogenes can grow throughout environmental samples were collected monthly throughout the shelf life of the product without temperature abuse. 1 year. L. monocytogenes was isolated from 3.8% of the raw Cold-smoked is a complex multistep salmon and 1.3% of the cold-smoked salmon. This study process that relies on effective sanitation controls to prevent used ribotyping of environmental samples to identify specific L. monocytogenes contamination (3). The heat applied during subtypes of L. monocytogenes and routes of contamination. The is not sufficient to eliminateL. monocytogenes (e.g., prevalence of L. monocytogenes in environmental samples from FDA recommends smoking not exceed 32°C (90°F)), and the four plants ranged from 0 to 30%. Drains and nonfood cold-smoked fish is considered a raw RTE food by FDA(20) . contact surfaces were identified as the most frequently Environmental sampling and molecular typing methods, contaminated areas. This study found that the four plants such as ribotyping, pulsed-field gel electrophoresis (PFGE), received similar raw ingredients and produced similar finished and, more recently, whole genome sequencing, have been products but varied in sanitation practices and procedures to used to compare isolates and subsequently identify routes prevent cross-contamination. Of the four plants, plant 1 had of contamination, which can then lead to the design of the highest prevalence of L. monocytogenes in environmental control measures (3, 16). A 1994 review article of 11 studies and finished product samples. The ribotyping data for plant reported that the overall prevalence of L. monocytogenes 1 showed that transient subtypes were constantly being on cold-smoked fish was approximately 10%, with the introduced into the plant by equipment, employees, and raw exception of two studies that found that 24 and 31% of the materials and also showed two persistent subtypes in the plant. samples contained L. monocytogenes (1). Concern over the One of the persistent subtypes of L. monocytogenes was found high prevalence of L. monocytogenes in cold-smoked fish in floor drains, on the gloves of an employee, and, ultimately, led researchers to track the presence of L. monocytogenes in in the finished product. In plant 2, althoughL. monocytogenes processing facilities using environmental sampling (8, 11, was not found in finished product, a persistent subtype was 19). A 1995 environmental sampling study of six cold- identified in the raw materials processing area that was also smoked salmon processing plants concluded that the surface found on a door handle adjacent to the finished product area, of frozen and fresh raw salmon was the primary source of indicating that cross-contamination from the raw materials L. monocytogenes contamination and that during processing area may have occurred. In plant 3, ribotyping data suggested this contamination was transferred to equipment, personnel, that a niche in the floor mat harbored a persistent subtype of and other surfaces, which subsequently became secondary L. monocytogenes, which subsequently contaminated finished sources of contamination (8). product. These results from plants 1, 2, and 3 showed the Other studies using environmental sampling with importance of effective cleaning and sanitization procedures ribotyping have found that insanitation is the most likely to eliminate niches where L. monocytogenes persists and the source of finished product contamination(11, 19). In 2000, importance of procedures to prevent cross-contamination. 512 environmental and 315 raw fish (whitefish, sablefish, Plant 4 was substantially different from the other three plants and salmon) samples were collected from two cold-smoked in that L. monocytogenes was not isolated from environmental fish processing plants (A and B) in the United States over or finished product samples; it was observed that the floors two 8-week periods (11). The results indicated that there and equipment had sufficient time to dry between processing was no significant statistical difference in the positive rate times, indicating the importance of a dry processing for L. monocytogenes in raw fish between plant A and plant environment. The study concluded thatL. monocytogenes B, and the overall prevalence of L. monocytogenes for all raw contamination in the processing environment was the most fish was 14.6%. However, the prevalence ofL. monocytogenes likely source of finished product contamination and that in environmental samples from plant A was 43.8% and from sanitation and L. monocytogenes control measures are critical.

March/April Food Protection Trends 185 Some processors use antimicrobial agents on raw ranged from >1 to 102 CFU/g, 7 ranged from >102 to 104 RTE fish fillets before cold smoking to reduce pathogenic CFU/g, and 2 ranged from >105 to 106 CFU/g. bacteria, including L. monocytogenes. Although antimicrobial A 2008 report on FDA’s strategy to combat listeriosis agents can be highly effective (5-log reduction) to sanitize indicated that the joint efforts of industry, government, clean food contact surfaces, seafood processors should not consumers, and academia led to a reduction in the incidence rely solely on antimicrobial agents to eliminate or reduce of listeriosis in the United States from seven to three cases Listeria on raw RTE salmon fillets. Whereas there is limited per million people between 1986 and 2006 (12). This report information in the scientific literature on the effectiveness of found that, from 1986 to 2006, fish was the most recalled antimicrobial agents in reducing Listeria on raw fish fillets, food for contamination with L. monocytogenes. three separate publications, one for acidified sodium chlorite More recently, a 2017 publication of a survey for L. mono- (ASC) and two for ozone on salmon fillets, all showed ≤1-log cytogenes in RTE foods suggested that the efforts of industry reduction (5, 6, 17). Su and Morrissey (17) evaluated the use and regulators to control L. monocytogenes have contributed of a wash of 50 ppm of ASC for 1 min on L. monocytogenes to a significant decrease in contaminated smoked seafood inoculated salmon fillets and found only a 0.52-log reduction (13). This survey provided the results of 27,389 food of L. monocytogenes; the authors suggested that the reduced samples collected weekly from supermarkets in California, efficacy of the ASC was probably due to reactions with Maryland, Connecticut, and Georgia from 2010 to 2013. organic substances, such as fish proteins. Crapo et al.(5) The survey foundL. monocytogenes in 2 (0.27%) of 745 evaluated the efficacy of ozonated water at 0.6 to 1.5 ppm to smoked seafood samples, which was significantly lower reduce Listeria innocua inoculated onto raw salmon fillets; than the 4.31% prevalence published in 2003 by Gombas they concluded that, whereas ozone can reduce bacterial et al. (9). Enumeration of L. monocytogenes from the two levels in pure culture, it was not effective at reducing positive samples showed a range of 9 to 250 CFU/g, which L. innocua (<1-log reduction) on salmon filets. The authors is less than the maximum concentration of 1.5 × 105 CFU/g also suggested that organic material, such as proteins, published by Gombas et al. (9). The authors suggested present on the surface of the fish fillets may react with the that this decrease in L. monocytogenes contamination was ozone in the water, reducing its effectiveness. Crowe et al. likely due to many factors, including industry practices and (6) evaluated the efficacy of ozone and found that the use regulatory policies. of three spray passes of 1 mg/L (1 ppm) aqueous ozone In January 2017, FDA issued the revised draft guidance generated a 1.17 ± 0.04-log reduction of Listeria inoculated for industry “Control of Listeria monocytogenes in Ready-to- on salmon fillets. Su and Morrissey(17) and Crapo et Eat Foods.” However, FDA continues to detect L. mono- al. (5) emphasized the importance of preventing cross- cytogenes in RTE food processing areas, leading to numer- contamination and of effective sanitation of food contact ous compliance actions. This paper offers insight into FDA surfaces as critical control measures for L. monocytogenes. inspections intended to identify controls and obtain data Effective control measures for L. monocytogenes are on the prevalence of L. monocytogenes within cold-smoked especially important for smoked seafood processors. salmon facilities. Although these inspections were conducted Smoked seafood was ranked as the fifth highest food in 2011, many of the observed insanitary conditions are category on a per serving basis for the risk of causing similar to recent noncompliant inspections of seafood listeriosis in the 2003 FDA, U.S. Department of facilities where FDA found L. monocytogenes in RTE food Agriculture risk assessment for L. monocytogenes (26). processing areas. The findings and recommendations in The risk assessment utilized data from 24 international this manuscript are relevant to all RTE food processors, and 6 domestic smoked seafood studies that showed that especially seafood processors. 13% of retail smoked seafood samples had detectable levels of L. monocytogenes. A preliminary determination MATERIALS AND METHODS of the risk assessment was a need for additional domestic FDA investigators conducted current good manufacturing data on the prevalence of L. monocytogenes in the United practice (CGMP) inspections and environmental sampling for States. Concurrently, a 2003 publication by Gombas et L. monocytogenes in 15 cold-smoked salmon facilities between al. (9) on RTE foods determined the presence and levels July and December 2011. The objectives of the inspections of L. monocytogenes in select retail foods from the United were to identify current operating practices and controls and States to estimate consumer exposure and support the to obtain data on the prevalence of L. monocytogenes. need for control strategies. This survey tested 31,705 CGMP inspections were conducted according to 21 CFR food samples from retail firms in Maryland and California Part 110 and focused on the potential for cross-contamination collected weekly from 2000 to 2001. L. monocytogenes was of finished product from equipment, food contact surfaces, found in 114 (4.31%) of 2,644 smoked seafood samples. food, and employees (25). Investigators discussed sanitation Enumeration of L. monocytogenes from the 114 positive procedures with management, reviewed their procedures if samples showed that 78 samples had ≥1 CFU/g, 27 written, and observed employees during cleaning and sanitiz-

186 Food Protection Trends March/April ing. Investigators also documented procedures intended to drains, etc. In addition to environmental samples, finished reduce the potential for contamination of finished product product samples were collected from 2 of the 15 facilities. from cross-contamination in the processing environment, Swab and finished product samples were analyzed following such as product flow, employee practices, and the use of the Bacteriological Analytical Manual method for the environmental sampling, to verify sanitation controls. The detection of Listeria (10). Samples found to contain Listeria number of employees (i.e., an indication of facility size), spp. were speciated to determine whether the isolates were employee training, and use of antimicrobial agents on L. monocytogenes or a nonpathogenic species of Listeria. fish prior to cold smoking were also recorded. This paper PFGE was performed on L. monocytogenes isolates using provides information on practices that were observed by a PulseNet’s standardized protocol for L. monocytogenes (4). different investigator in each individual facility. This does not allow us to compare practices done in one facility to those RESULTS in another facility to determine the significance of different Environmental samples practices to minimize cross-contamination. Table 1 summarizes the environmental sampling results Environmental samples (i.e., swabs) were collected and the number of employees in each of the 15 cold-smoked from food contact surfaces and nonfood contact surfaces fish processing facilities inspected and sampled by FDA. The in close proximity to food and food contact surfaces (24). facilities varied in size from the smallest, with one part-time Approximately 50 to 100 swabs were collected at each facility. employee, to the largest, with 180 full-time employees and The number of swabs collected was based on the size of the with additional employees hired seasonally. For discussion facility. Environmental samples were taken several hours purposes, these 15 facilities were placed into four groups into production by swabbing surfaces such as food contact based on their environmental sample results. surfaces, smoking equipment, totes, cracks in the floor, floor

TABLE 1. FDA environmental swab results for Listeria spp. and L. monocytogenes positives and number of employees for each of the 15 cold smoked fish processing facilities

Nonpathogenic Number of Employees Listeria spp. L. monocytogenes Facility Group1 Species of Listeria Positives Positives Positives Full-time Part-time A 1 0 0 0 35 25 B 1 0 0 0 6 4 C 1 0 0 0 22 0 D 2 1 0 1 100 0 E 2 1 0 1 0 1 F 2 1 0 1 18 0 G 2 3 0 3 180 Seasonal H 2 1 0 1 4 0 I 3 2 1 1 Not recorded J 3 1 1 0 6 0 K 3 1 1 0 90 0 L 3 1 1 0 4 0 M 3 1 1 0 Not recorded N 4 23 11 12 2 0 O 4 12 10 2 35 0

1. Group designation was based on FDA environmental sampling results for Listeria spp. and L. monocytogenes.

March/April Food Protection Trends 187 TABLE 2. Sanitation practices used in each of the three facilities in Group 1 that had no Listeria spp. detected in FDA environmental samples

Written Flow of processing Facility sanitation Cleaning and sanitation practices Foot-baths to prevent cross- procedures contamination

Full cleaning and sanitizing of food contact and environmental surfaces throughout production and at A Yes Not observed Yes end of each production day with varying cleaning and sanitizing chemicals.

End of production debris removed, and food contact surfaces scrubbed with brushes, warm water and B Yes soap. Cleaning and sanitizing conducted again before Not observed Not observed production. Utensils and removeable parts cleaned and sanitized. Floor drains foamed three times a week.

Entire area cleaned every four hours with foaming system. Equipment disassembled washed, rinsed and sanitized. C Not observed Yes Yes Fixed equipment rinsed to remove debris, foamed, scrubbed, hot water rinsed and sanitized.

Observations corrective action, product was no longer stored under the Group 1, sanitation controls. All three facilities in group condenser and management contacted a contractor to assess 1 (A, B, and C), which had no Listeria spp. detected in the the cooling unit. The processor promised further corrective environmental samples, routinely conducted environmental actions. The investigator did not describe the final disposition and finished product testing. All three facilities also provided of the product; however, the adulterated product should have frequent training for their employees on CGMP and food been destroyed or diverted to a nonfood use. safety. Table 2 summarizes additional sanitation practices Group 2, environmental sample locations. The environmental used in these three facilities. sample results from the five facilities in group 2 (D, E, F, G, Facility A had the most practices to minimize cross- and H) showed that Listeria spp. were detected without contamination among the three facilities in group 1. The flow L. monocytogenes; this indicated that nonpathogenic stains of raw materials, finished product, and employees through of Listeria were isolated from sites that were in close contact the facility was controlled by sanitation procedures. The or direct contact with product. These sites consisted of a fish raw material receiving area, rinsing and salting area, curing brine rack, pin bone remover belt, handles of dry cure mixer, cooler, and smoking room were separate from the smoked a basin of a stainless steel sink used for rinsing fish, and a fish blast chiller and smoked fish processing area. Employees crack in a processing room floor. were restricted from moving among the various processing Group 2, sanitation controls. Facility D had the most areas. Employees wore facility-provided boots, hair nets, documented practices to minimize cross-contamination and gloves in processing areas. To verify sanitation controls, among the 15 facilities inspected and sampled. The cleaning facility A collected environmental samples weekly from six and sanitizing program included written procedures for food contact surfaces and analyzed them for Listeria spp. equipment, coolers, tables, floors, walls, ceilings, drains, The investigator’s review of the facility’s own environmental condensers, drip pans, etc. with monitoring records for results showed no positives for Listeria spp. Facility B cleaned specific daily, weekly, and monthly tasks. The food contact by scrubbing equipment with brushes and detergent and surfaces in the processing rooms were wiped down with 200 then by sanitizing both before and after production. ppm of quaternary ammonia before employee breaks and Group 1, insanitary conditions. Facility C had condensate, lunch, with a full wet cleaning and sanitizing after production. a common source of Listeria, dripping from a cooling unit Coolers were emptied and then cleaned and sanitized weekly onto exposed RTE finished product of cold-smoked salmon, (floors and walls) and monthly (ceilings and evaporators) which is an egregious sanitation deficiency. As an immediate by a designated cleaning crew. Employees wore clean lab

188 Food Protection Trends March/April coats, designated boots, hair nets, sleeve guards, and gloves. Facility F had condensate dripping from the condenser Hand wash and boot dip stations were located adjacent to in the raw cooler in four different areas, creating four pools the employee changing room. Additional boot dips and on the floor near uncovered salmon (not smoked). Also, in hand dips were strategically located throughout the facility, the finished product cooler there were two pools of standing and the concentration of the sanitizer was tested by the water. Pools of water and condensate on floors have the quality control manager. A nonalkaline detergent was used in potential to be contaminated with Listeria and to cross- addition to an alkaline chlorinated foaming system. Peracetic contaminate wheels of moveable equipment, such as carts acid and a quaternary ammonium were used to sanitize and forklifts, which can spread contamination throughout the equipment. The flow of processing was done in one direction, facility. Additionally, when carts are pushed through pools from ingredients to finished product, that minimized of liquid on floors there is the potential for splash to cross- potential cross-contamination of finished product. There was contaminate any food or items being conveyed on the carts. a physical separation of the finished product processing area In this facility, the floors in the coolers and processing areas from the other areas within the facility. Employee movement were cracked and pitted. Floors in disrepair have the potential was restricted by using area-specific designated employees. to be difficult to clean and can create niches that harbor Moveable equipment circulation was restricted by using area- Listeria. Employees in facility F were observed wheeling specific designated carts, pallet jacks, and trash receptacles. garbage cans from the finished product processing area Finished products were stored in a separate cooler. The outside to the dumpster and then returning through multiple facility also had employee training upon hire and annually, processing areas to the finished product processing area with topics including GMP, hazard analysis and critical without cleaning and sanitizing the equipment. control point, chemicals, hand washing, food safety, Listeria Facility H had a significant number of flies in the controls, etc. Environmental samples were collected weekly processing room, and six flies were observed walking on from 10 sites on food contact surfaces and nonfood contact a preparation table. Flies can carry pathogens and cross- surfaces from high-risk areas. The environmental samples contaminate food, food contact surfaces, and equipment. were analyzed for Listeria spp. by the quality control manager. Group 3, environmental sample locations. Environmental The investigator’s review of the facility’s own environmental sample results from the five facilities in group 3 (I, J, K, L, results for the previous 6-month period showed that Listeria and M), which each had one swab with L. monocytogenes spp.-positive results were found on a wire cart, floor drain, detected, showed that L. monocytogenes was isolated from and slicer. The facility’s documented corrective actions for drains, floors, and equipment in direct contact with the floor the wire cart and floor drain positives included aggressive in four of the five facilities. In the fifth facility,L. monocy- cleaning and sanitizing with disassembly and negative results togenes was isolated from a wooden cutting board used as a for retesting of the affected and surrounding areas. The wire food contact surface during packing of smoked fillets. This cart was eventually removed from service because it was facility (L) was re-inspected by FDA to evaluate the imple- determined by the facility to be hard to clean. The slicer was mented corrective actions for the detection of L. mono- removed from service, aggressively cleaned and sanitized, cytogenes on a food contact surface (i.e., the wooden cutting and then retested with 10 additional swabs. Two Listeria spp. board). The wooden cutting board had been discarded. Also, positives were also found by the firm from the area of (i) top during the re-inspection, FDA collected 10 subsamples of and bottom sprockets on the toothed feed belt (lattice area) cold-smoked salmon that were negative for L. monocytogenes and (ii) platform. The lattice area was taken apart for more and the facility was found to be in compliance with CGMP intense cleaning and sanitizing followed by retesting with six according to 21 CFR Part 110. swabs that were negative. The next day the lattice area of the Group 3, insanitary conditions. In group 3, four of the same slicer was sampled (six swabs) after soaking the entire five facilities did not have any significant objectionable unit in 200 ppm of quaternary ammonia solution, and five observations during the inspection. The fifth facility (I) had swabs were positive. The lattice area was heat treated at 150°F three CGMP deficiencies related to plumbing, floors, walls, (65.6°C) for 2 h, and the retesting results were negative. and ceilings that warranted voluntary action. Group 2, insanitary conditions. Facility E had condensate Group 4, facility N environmental sample locations. Results dripping onto a cooler floor near uncovered cooling cold- of environmental sampling from facility N showed that smoked fish. A hose with an affixed nozzle used to clean Listeria spp. were detected in 23 swabs from the 78 collected. food contact surfaces was stored on a wet floor, allowing After further testing of the 23Listeria spp. isolates, 11 were the potential for cross-contamination from the floor to food found to be L. monocytogenes and 12 were nonpathogenic contact surfaces. Dried food debris was on the blades of the Listeria species. Nonpathogenic Listeria species were isolated skinner and inside the grinder, indicating that the equipment from four legs of processing tables, two legs of brine tanks, was not properly cleaned after use. Conveyor belts on the two brine tank drains, the bottom of a brine tank, the front of finished product slicer were frayed and cracked, possibly a processing table, a strip curtain, and a metal basket on the making them difficult to clean and sanitize. floor used to hold fish waste.L. monocytogenes was found in a

March/April Food Protection Trends 189 brine tank drain, in the bottom of a brine tank, on two legs of the finished product processing floor for several feet while processing tables, in five areas on the floor, and in two floor leaving on break. drains in the processing areas. Group 4, facility O first re-inspection. Facility O was re- Group 4, facility N insanitary conditions. Totes used to keep inspected due to the detection of L. monocytogenes on a fish immersed in the brine solution were soiled and stored food contact surface (the bottom rack of a cart in the brine on the wet floor. Totes constructed with holes in the bottom cooler holding exposed salmon fillets on above racks). used for holding fish were placed on cardboard that had been During the re-inspection, the firm provided information on previously stepped on by employees. The fans used to dry their corrective actions for the previous inspection, which the fish after brining were visibly dirty. There was rust on consisted of cleaning and documentation of sanitation the ceiling in the walk-in cooler above exposed stored fish. improvements. Additionally, the FDA investigators observed A box of salt used during brining of the fish was stored on top employees handling processing clothing properly without of a floor drain. All of these sanitation issues were corrected contacting the floor, and the cooler had been reorganized before the inspection was completed. The firm did not have to prevent product from being stored under the cooler unit any procedures for environmental sampling. where dripping condensation was previously observed. The Group 4, facility O environmental sample locations (initial firm also replaced their totes with disposable totes, replaced inspection). Environmental sampling results from facility drain covers in the area, added additional foot dips, O showed that Listeria spp. were detected in 12 swabs from sanitized brining utensils, and started periodically spraying the 78 collected. After further testing of the 12Listeria step ladders with chlorine. spp. isolates, 10 were found to be L. monocytogenes and 2 Additional environmental sampling conducted by FDA were nonpathogenic Listeria species. L. monocytogenes was showed that L. monocytogenes was detected in 8 swabs of the detected on (i) a paddle used to agitate thawing fish, (ii) the 96 analyzed. L. monocytogenes was isolated from (i) a rubber bottom rack of a cart in the brine cooler, (iii) a brine tank grip area on the top handle of a step ladder in the brine area, drain spout, (iv) the exterior door handle on the smoker (ii) a taped area on the top handle of a step ladder in the cooler, (v) the rubber strip on the interior of the brine cooler brine area, (iii) a rubber strip on the base of the brining room door, (vi) the rubber grip areas on the top handles of two door, (iv) a wheel of a pallet jack, (v) a paddle with rivets step ladders (brining area), (vii) the front and (viii) rear used to stir brine, (vi) an in-floor scale in the fillet area, forklift wheels (thaw area), (ix) the exterior strip along the (vii) the front wheel of a pallet jack in the RTE food base of the thaw room door, and (x) a rope attached to the processing area, and (viii) rack surfaces in the finishing cooler. thaw room door used to lower the roll-up door. Finished product samples collected by FDA were found to Group 4, facility O sanitation controls (initial inspection). be contaminated by L. monocytogenes. The firm voluntarily Facility O had ozonated water for thawing fish, quaternary recalled the product. PFGE conducted by FDA showed that ammonium hand dips, fogs for equipment, foot dips, foaming the L. monocytogenes isolated during the initial inspection floor sprays, peracetic floor wash, quarterly employee was indistinguishable from the isolate obtained during training, an environmental testing plan for Listeria spp., the first re-inspection. This indicates that the facility’s and finished product testing forL. monocytogenes. The cleaning, sanitizing, and corrective actions were not investigator’s review of the facility’s own routine weekly sufficient to eliminate the pathogen and a resident strain environmental sampling over a 10-month period prior to the of L. monocytogenes was established in the facility. These inspection showed that Listeria spp. were detected on two results also suggest that the rubber grip and taped areas on separate sampling dates from the floor in the brining room the step ladders may have been difficult to clean surfaces and the floor near a freezer. The firm’s corrective actions that created niches for the harborage of L. monocytogenes included cleaning with chlorine solution, steam cleaning, and and that the firm’s corrective action of periodically spraying sanitizing with quaternary ammonia solution, followed by the the step ladders with chlorine did not eliminate the collection of additional swabs over three consecutive days. pathogen. The PFGE conducted by FDA also showed that The firm’s results showed thatListeria spp. were not detected L. monocytogenes isolated from the finished product was in the swabs taken after this intense cleaning and sanitizing. indistinguishable from the environmental samples. This Group 4, facility O insanitary conditions (initial inspection). indicates that the contamination most likely occurred from During the inspection, investigators observed poor employee cross-contamination within the facility. practices. For example, employees sprayed and cleaned Group 4, facility O second re-inspection. A second re-inspec- brining bins in a manner that generated splash and spray, tion was conducted because FDA detected L. monocytogenes which contacted the floor near fish that was ready to enter the in environmental and finished product samples during the smoker; this created the potential for cross-contamination first re-inspection. Once again, FDA evaluated CGMPs, col- from the floor. Carts of packaged finished product were lected environmental and finished product samples, verified stored under cooling units that were dripping condensation. sanitation and process changes since the first re-inspection, An employee was observed dragging their smock across and reviewed the firm’s environmental testing records. The

190 Food Protection Trends March/April TABLE 3. Use of antimicrobial agents on fish and Listeria spp. environmental sampling results for eight of the 15 cold smoked fish processing facilities

Safe and suitable use of Listeria spp. Facility Group Antimicrobial agents used on fish antimicrobial positives agents

A 1 Chlorine dioxide rinse 25 ppm No 0 B 1 Peracetic acid <160 ppm and Chlorine dioxide 40–50 ppm, separately No 0 D 2 ASC rinse 40–50 ppm for 5 minutes with subsequent water rinse Yes 1 E 2 Quaternary ammonium 150 ppm No 1 G 2 Fillets thawed in ASC at 2–5 ppm with subsequent water rinse Yes 3 L 3 ASC rinse 100 ppm for <15 minutes, no subsequent water rinse No 1 M 3 ASC rinse 40–50 ppm for 30–60 minutes, no subsequent water rinse No 1 O 4 Ozonated water rinse Yes 12

(ASC) Acidified Sodium Chlorite

firm documented the disposal of the difficult to clean step Of the 15 facilities inspected, eight used antimicrobial ladders and the riveted paddle used in brining that were agents on fish(Table 3). Of these eight facilities, five (A, B, found to be contaminated with L. monocytogenes by FDA. E, L, and M) used antimicrobial agents in a manner that did The firm continued their own environmental sampling not conform with authorized uses. Facility A inappropriately and began holding finished product until test results were used a 25-ppm chlorine dioxide rinse that exceeds the limit of negative for L. monocytogenes. The firm’s own environmental 3 ppm listed in effective food contact notifications for chlo- testing results were all negative for L. monocytogenes. The rine dioxide use on seafood. Facility B inappropriately used a environmental and finished product samples collected by rinse of <160 ppm of peracetic acid followed by a 40- to 50- FDA during the second re-inspection were negative for ppm chlorine dioxide rinse on fish. There is no food additive Listeria spp., and the inspection was classified as voluntary regulation nor a food contact notification allowing the use action indicated. of peracetic acid by itself on seafood. Additionally, the use of 40 to 50 ppm of chlorine dioxide exceeds the limit of 3 ppm Use of antimicrobial agents on fish listed in effective food contact notifications for chlorine dioxide Any substance used on fish needs to be safe and suitable use on seafood. Facility E inappropriately used 150 ppm of for use and in compliance under the applicable regulation. quaternary ammonium on fish. Quaternary ammonium is The use of antimicrobial agents on fish before smoking approved for use on processing equipment and food contact was documented in some of the facilities during the surfaces, but not as an antimicrobial agent on seafood. inspections of the 15 facilities (Table 3). It is important Four facilities (D, G, L, and M) used ASC as an antimicro- for seafood processing facilities to comply with FDA bial agent in water to rinse fish. ASC can be used to rinse regulations (including specifications and conditions of seafood in accordance with current industry standards at a safe use) or to comply with the conditions of safe use of a concentration of 40 to 50 ppm of sodium chlorite when generally recognized as safe substance when using permitted followed by a potable water rinse prior to consumption for antimicrobial agents in processing seafood. For example, fish that is intended to be consumed raw (e.g., cold-smoked there are health concerns in regard to possible residues fish), per 21 CFR 173.325 (d)(1). Of the four facilities, of chlorite and chlorate on treated seafood, or possible two (D and G) used ASC in accordance with food additive mutagenic and carcinogenic reaction products that may result regulations under 21 CFR 173. ASC use was inappropriate from misuse of chlorine dioxide and ASC directly on food in facilities L and M because treated fish was not subjected to (18). FDA considers food treated with antimicrobial agents a water rinse as required for cold-smoked fish. Additionally, in an unauthorized manner to be in violation of the law and facility L used 100 ppm of ASC, which exceeds the concen- potentially unsafe. trations of 40 to 50 ppm listed in 21 CFR 173.325 (d)(1).

March/April Food Protection Trends 191 Facility O used ozonated water to rinse fish. Ozone Environmental sampling should be conducted to verify in water can be used as an antimicrobial agent on food, the effectiveness of cleaning and sanitizing and to identify including seafood, in accordance with the regulations under and eliminate harborages of L. monocytogenes (15). L. mono- 21 CFR Part 173.368. cytogenes may not be eliminated by routine sanitization because it is capable of producing biofilms and persisting DISCUSSION in processing facilities by establishing harborages and Facilities N and O were found by FDA to have a higher niches (3). Processors should sample after several hours prevalence of L. monocytogenes (11 and 10 positives, of production to allow harborages to loosen and L. mono- respectively). Facilities N and O were also found to have cytogenes to be dislodged (15). Sampling immediately several insanitation observations, such as uncontrolled following cleaning and sanitizing can result in consistently dripping condensate, cross-contamination, and an ineffective negative results and a failure to identify harborages (15). or no environmental sampling program. Processors should also test environmental samples for the During the initial inspection, facility O had carts of presence of Listeria spp. Although L. monocytogenes is the packaged finished product stored under cooling units that most concerning species of Listeria that causes human were dripping condensation, a common source of Listeria. illness, the presence of Listeria spp. is often used to identify Sanitation monitoring for the prevention of adulteration of conditions conducive to the presence and growth of L. mono- food, food packing material, and food contact surfaces by cytogenes (3, 15). FDA’s draft guidance for industry condensate is required in 21 CFR 123.11 (b)(5). Corrective “Control of Listeria monocytogenes in Ready-to-Eat Foods” actions need to ensure that no product enters commerce recommends that processors test environmental samples for that is either injurious to health or is otherwise adulterated, the presence of Listeria spp. without determining whether in addition to correcting the cause of the problem (e.g., L. monocytogenes is present (23). Testing for Listeria spp. elimination of condensation, repair or maintenance of detects both L. monocytogenes and the more common condenser). The cooler was reorganized to prevent product nonpathogenic species (15, 23). To ensure that Listeria from being stored under the cooler unit where dripping control measures eliminate harborages of Listeria, processors condensation was previously observed. However, although should conduct robust environmental sampling with a “seek moving product away from dripping condensate is an and destroy” mentality (23). The draft guidance recommends acceptable immediate and preliminary correction, this that processors determine the appropriate number of correction does not correct the cause of the uncontrolled sampling sites based on the size of the plant, plant features, dripping condensate. product flow, characteristics of the RTE food, the processing Facility O was found by FDA to have L. monocytogenes in methods used to produce the food, and any previous 10 environmental samples in the initial inspection and in sampling results. Even the smallest processors should collect eight environmental samples and finished product in the first weekly samples after several hours of production from at re-inspection. The environmental sampling results showed least five food contact surfaces and five nonfood contact that the firm had inadequate cleaning and sanitizing because surfaces on each RTE production line of food that supports the step ladders used in the brining area to place fish on trays the growth of L. monocytogenes (e.g., cold-smoked salmon). into racks for the smoker that were found positive during Additionally, the draft guidance recommends that all food the initial inspection were still contaminated with L. mono- contact surfaces be tested at least once each month and all cytogenes in the first re-inspection. The PFGE conducted by nonfood contact surfaces at least once each quarter. With FDA showed that a resident strain of L. monocytogenes was this amount of sampling, processors should expect to detect established in the facility and that L. monocytogenes isolated Listeria spp. occasionally in a few environmental samples from the finished product most likely occurred from cross- and be ready with appropriate corrective actions based on contamination within the facility. Although facility O had location. several sanitation controls in place, it appears that the firm’s Sampling food contact surfaces and taking appropriate use of ozonated water on raw product and chemicals on corrective actions is critical due to the risk of contaminating food contact surfaces and floors did not compensate for the food. During the initial inspection, facility O was found by insanitary practices. FDA to have L. monocytogenes on a food contact surface Facility O appeared to have an ineffective environmental (the bottom rack of a cart in the brine cooler holding sampling plan. In view of FDA detecting 10 L. monocytogenes exposed salmon fillets on above racks). The firm’s corrective positives in one sampling and the firm detecting only two actions included cleaning and documentation of sanitation positives over a 10-month period, it appears that the firm’s improvements. When L. monocytogenes is detected on food environmental sampling was ineffective. Firms should contact surfaces, such as racks, carts, utensils, brining tanks, routinely examine their sampling frequency and timing and cutting boards, FDA recommends that processors of collection along with the number and location of samples implement corrective actions that address both the facility collected. and the food (23). Facility-related corrective actions should

192 Food Protection Trends March/April include intensified cleaning and sanitizing with disassembly surfaces during the same sampling period. This is because of equipment, intensified environmental sampling of affected the routine sanitation procedures are likely inadequate, and and surrounding areas, and correction of any conditions Listeria spp. may have become established in one or more or procedures that led to the contamination. Food-related harborages (23). corrective actions should include segregation and holding of any food that may have had contact with the contaminated CONCLUSION food contact surface. Processors should then decide what to The inspections showed a low prevalence ofL. mono- do with the affected food. FDA recommends the product be cytogenes (≤1 positive) in each of the 13 of 15 cold-smoked reprocessed with a listericidal control measure (e.g., a cook salmon processing facilities sampled by FDA. Two facilities that eliminates L. monocytogenes), destroyed, or diverted to a (N and O) had a higher prevalence of L. monocytogenes (11 and use in which it will not be consumed by humans or animals. 10 positives, respectively). Facility N had 11 L. monocytogenes If product has entered commerce, processors should initiate and 12 nonpathogenic species of Listeria positives, mostly from a recall. the floor, floor drains, and equipment in contact with the floor. Facility N was found by FDA to have L. monocytogenes The facility had no environmental sampling program, and their in a brine tank drain, on the bottom of a brine tank, on two most significant insanitary conditions were poor employee legs of processing tables, in five areas on the floor, and in two practices that likely led to Listeria spp. cross-contamination floor drains in the processing areas. These environmental from the floor to equipment. Facility O had gross insanitation, sampling results indicate that the nonfood contact surfaces poor employee practices, difficult to clean surfaces, and of floor, floor drains, and equipment in contact with the floor inadequate cleaning and sanitizing that likely contributed to were potential sources of L. monocytogenes contamination. cross-contamination, resulting in the adulteration of cold- Nonfood contact surfaces can be a potential source of smoked salmon from a resident strain of L. monocytogenes. contamination, and having adequate corrective actions is In addition, the facility’s use of antimicrobial agents on raw important. Drains are a composite of what can be found in product, food contact surfaces, and floors did not compensate the room and are a hospitable environment for Listeria (15). for the insanitary practices observed during the inspection. When Listeria is detected in drains, processors should correct The results also showed that three facilities (A, B, and any conditions that led to the contamination, including C), where there was no detection of Listeria spp., routinely evaluation of drain design and function; conduct additional conducted environmental and finished product testing environmental sampling; clean and sanitize the area; and and had frequent employee training in addition to routine ensure that procedures are in place to prevent the spread cleaning and sanitizing. Overall, the results show that to of contamination (23). Although surfaces such as floors, minimize the potential for L. monocytogenes contamination, drains, and wheels of a pallet jack or cart are not food contact it is necessary to have comprehensive practices that include surfaces, these surfaces are important because insanitary sanitation that prevents contamination of food and food actions by employees can spread the contamination to food contact surfaces, sanitary employee practices, appropriate contact surfaces or food (23). For example, contamination corrective actions, and environmental sampling to identify on the floor could be transferred to food contact surfaces and eliminate niches where L. monocytogenes can persist. by water splash and spray or by placing items, such as boxes and hoses, on food contact surfaces after they have been in ACKNOWLEDGMENTS direct contact with the floor. WhenListeria spp. are detected The authors are appreciative of FDA investigators that on nonfood contact surfaces that are in close proximity conducted the inspections and collected environmental to food and food contact surfaces, FDA recommends samples. The authors also thank the following FDA, Center that processors implement corrective actions that include for Food Safety and Applied Nutrition employees for their cleaning and sanitizing, environmental sampling of affected thoughtful review of the manuscript: Jenny Scott, Senior and surrounding areas, and correction of any conditions or Advisor to the Director, Office of Food Safety; Mickey procedures that led to the contamination (23). Additionally, Parish, Senior Science Advisor, Office of the Center Director, FDA recommends that sanitation procedures be intensified and Lane Highbarger and Molly Harry, Consumer Safety when Listeria spp. are detected on several nonfood contact Officers, Office of Food Additive Safety. surfaces that are in close proximity to food and food contact

REFERENCES 1. Ben Embarek, P. K. 1994. Presence, detection 3. Busta, F. F., G. E. Bledsoe, G. J. Flick, 4. Centers for Disease Control and Prevention. and growth of Listeria monocytogenes in L. Gram, D. Herman, M. L. Jahncke, and 2013. Standard operating procedure for : a review. Int. J. Food Microbiol. D. R. Ward. 2001. Processing parameters PulseNet PFGE of Listeria monocytogenes. 23:17–34. needed to control pathogens in cold-smoked Available at: https://www.cdc.gov/pulsenet/ fish.J. Food Sci. 66(Suppl.):S1055–S1134. pathogens/pfge.html. Accessed 15 May 2020. 2. Bennion, J. R., F. Sorvillo, M. E. Wise, Available at: https://www.fda.gov/Food/ S. Krishna, and L. Mascola. 2008. Decreasing FoodborneIllnessContaminants/ucm545157. listeriosis mortality in the United States, htm. Accessed 15 May 2020. 1990–2005. Clin. Infect. Dis. 47:867–874.

March/April Food Protection Trends 193 5. Crapo, C., B. Himelbloom, S. Vitt, and 14. Rocourt, J., C. Jacquet, and A. Reilly. 2000. 22. U.S. Food and Drug Administration. L. Pedersen. 2004. Ozone efficacy as a Epidemiology of human listeriosis and 2012. Listeria monocytogenes. In Bad bug bactericide in seafood processing. J. Aquat. seafoods. Int. J. Food Microbiol. 62:197–209. book: handbook of foodborne pathogenic Food Prod. Technol. 13:111–123. microorganisms and natural toxins, 2nd ed. 15. Scott, V. N., M. Wiedmann, D. Hicks, Available at https://www.fda.gov/food/ 6. Crowe, K., D. Skonberg, A. Bushway, and R. Collette, M. L. Jahncke, and K. Gall. foodborne-pathogens/bad-bug-book-second- S. Baxter. 2012. Application of ozone sprays 2005. Guidelines for Listeria testing of edition. Accessed 15 May 2020. as a strategy to improve the microbial safety environmental, raw product and finished and quality of salmon fillets.Food Control product samples in smoked seafood 23. U.S. Food and Drug Administration. 25:464–468. processing facilities. Food Prot. Trends 2017. Draft guidance for industry: control 25:23–34. of Listeria monocytogenes in ready-to-eat 7. Dalgaard, P., and L. V. Jorgensen. 1998. foods. Available at: https://www.fda.gov/ Predicted and observed growth of Listeria 16. Stasiewicz, M. J., H. F. Oliver, M. Wiedmann, regulatory-information/search-fda-guidance- monocytogenes in seafood challenge tests in and H. C. den Bakker. 2015. Whole-genome documents/draft-guidance-industry-control- naturally contaminated cold-smoked salmon. sequencing allows for improved identification listeria-monocytogenes-ready-eat-foods. Int. J. Food Microbiol. 40:105–115. of persistent Listeria monocytogenes in food- Accessed 15 May 2020. associated environments. Appl. Environ. 8. Eklund, M. W., F. T. Poysky, R. N. Paranjpye, Microbiol. 81:6024–6037. 24. U.S. Food and Drug Administration. 2020. L. C. Lashbrook, M. E. Peterson, and G. A. Pelroy. Sampling: collection technique, subchap. 4.3. 1995. Incidence and sources of Listeria mono- 17. Su, Y., and M. Morrissey. 2003. Reducing In Investigations operations manual. Available cytogenes in cold-smoked products and levels of Listeria monocytogenes contamination at: https://www.fda.gov/inspections- processing plants. J. Food Prot. 58:502–508. on raw salmon with acidified sodium chlorite. compliance-enforcement-and-criminal- J. Food Prot. 66:812–818. 9. Gombas, D. E., Y. Chen, R. S. Clavero, investigations/inspection-references/ and V. N. Scott. 2003. Survey ofListeria 18. Taylor, J. B., and D. W. Wohlers. 2004. investigations-operations-manual. Accessed monocytogenes in ready-to-eat foods. J. Food Toxicological profile for chlorine dioxide 2 October 2020. Prot. 66:559–569. and chlorite. U.S. Environmental Protection 25. U.S. Food and Drug Administration. Agency, Washington, D.C. Available at: 10. Hitchins, A. D., K. Jinneman, and Y. Chen. 2020. Establishment inspections: food, https://stacks.cdc.gov/view/cdc/37580. 2016. Detection and enumeration of subchap. 5.4. In Investigations operations Accessed 15 May 2020. Listeria monocytogenes in foods, chap. 10. In manual. Available at https://www.fda. Bacteriological analytical manual. Available 19. Thimothe, J., K. Kerr Nightingale, K. Gall, gov/inspections-compliance-enforcement- at: https://www.fda.gov/food/laboratory- V.N. Scott, and M. Wiedmann. 2004. and-criminal-investigations/inspection- methods-food/bam-detection-and-enumera- Tracking of Listeria monocytogenes in references/investigations-operations-manual. tion-listeria-monocytogenes. Accessed smoked fish processing plants. J. Food Prot. Accessed 2 October 2020. 15 May 2020. 67:328–341. 26. U.S. Food and Drug Administration and 11. Hoffman, A. D., K. L. Gall, D. M. Norton, and 20. U.S. Food and Drug Administration. U.S. Department of Agriculture. 2003. M. Wiedmann. 2003. Listeria monocytogenes 2011. Fish and fishery products hazards Quantitative assessment of the relative risk contamination patterns for the smoked fish and controls guidance, chap. 13. Avail- to public health from foodborne Listeria processing environment and for raw fish. able at https://www.fda.gov/food/ monocytogenes among selected categories J. Food Prot. 66:52–60. seafood-guidance-documents-regulatory- of ready-to-eat foods. Available at: https:// information/fish-and-fishery-products- www.fda.gov/food/cfsan-risk-safety- 12. Klontz, K. C., P. V. McCarthy, A. R. Datta, hazards-and-controls-guidance-4th-edition. assessments/quantitative-assessment- J. O. Lee, D. W. K. Acheson, and R. E. Accessed 15 May 2020. relative-risk-public-health-foodborne-listeria- Brackett. 2008. Role of the U.S. Food and monocytogenes-among-selected. Accessed Drug Administration in the regulatory 21. U.S. Food and Drug Administration. 15 May 2020. management of human listeriosis in the 2011. Fish and fishery products hazards United States. J. Food Prot. 71:1277–1286. and controls guidance. Appendix 4. Available at https://www.fda.gov/food/ 13. Luchansky, J. B., Y. Chen, A. C. Porto-Fett, seafood-guidance-documents-regulatory- R. Pouillot, B. A. Shoyer, R. Johnson- information/fish-and-fishery-products- DeRycke, D. R. Eblen, K. Hoelzer, W. K. hazards-and-controls-guidance-4th-edition. Shaw, Jr., J. M. van Doren, and M. Catlin. Accessed 15 May 2020. 2017. Survey for Listeria monocytogenes

in and on ready-to-eat foods from retail

establishments in the United States (2010

through 2013): assessing potential changes

of pathogen prevalence and levels in a decade.

J. Food Prot. 80:903–921.

194 Food Protection Trends March/April